Near-Earth asteroid 162173 (1999 JU3) is the primary target of the Hayabusa-2 sample return mission, and a potential target of the Marco Polo sample return mission. Earth-based studies of this object are fundamental to these missions. We present a mid-infrared spectrum (5-38 microns) of 1999 JU3 obtained with NASA's Spitzer Space Telescope in May 2008. These observations place new constraints on the surface properties of this asteroid. To fit our spectrum we used the near-Earth asteroid thermal model (NEATM) and the more complex thermophysical model (TPM). However, the position of the spin-pole, which is uncertain, is a crucial input parameter for constraining the thermal inertia with the TPM; hence, we consider two pole orientations. In the extreme case of an equatorial retrograde geometry we derive a lower limit to the thermal inertia of 150 J/m^2/K/s^0.5. If we adopt the pole orientation of Abe et al. (2008a) our best-fit thermal model yields a value for the thermal inertia of 700+/-200 J/m^2/K/s^0.5 and even higher values are allowed by the uncertainty in the spectral shape due to the absolute flux calibration. The lower limit to the thermal inertia, which is unlikely but possible, would be consistent with a fine regolith similar to wthat is found for asteroid 433 Eros. However, the thermal inertia is expected to be higher, possibly similar to or greater than that on asteroid 25143 Itokawa. Accurately determining the spin-pole of asteroid 162173 will narrow the range of possible values for its thermal inertia.